Building A Bear

This summer we radically improved KURB's search technology in ways I didn't think possible, though it's a reminder that this isn't the first time the project has been fundamentally rebuilt. Two things have never changed: the philosophy driving us to make the secondhand designer ecosystem as accessible as possible, and our neurotic conviction to build a sick tool for the community. Everything else about KURB has evolved, repeatedly, since day one. So I thought it'd be the right time to lay out that history and capture what was in the CPU-water-cooler at each stage.

The KURB we know today was cast from frustration with finding clothing that met my grandiose sartorial inclinations on a budget. In short: high ambition with limited resources, a pairing that typically makes any problem nightmarish. Miraculously, the secondhand designer market was exactly the key to solving that consumer struggle. But these markets had a problem of their own: accessibility and fragmentation. Even after narrowing the search to select markets and optimizing the process, actively surveying them so nothing coveted slipped by was its own hellish toil. To make matters worse, I'd voluntarily flung myself into deeper levels of pain as a (now retired) avid Carpe Diem (and their ilk) collector. Given enough time in these communities, it became clear this was a shared suffering, more like an asylum of chronically online dark fashion secondhand consumers than a solitary cell. 

Cue Threadbare, our first foray into building a freely accessible community tool that solved all of the above, starting mid-2022. It's what our Discord server today started as, and the technology is largely the same. It was a system that searched the most popular marketplaces and chains (around 50ish at the time) and fed a Discord bot that posted new finds on a schedule into delegated channels. That meant Discord users got notified the moment new listings from a given pool of brands hit the market. Systems like Threadbare are common in fashion servers now, but at the time it was the most dedicated solution out there. I'd add features on intuition and fix bugs based on community feedback, but I wasn't actively developing the project so much as focusing on the technology itself. Still, I grew increasingly frustrated with the limitations of Discord and having to continually invent creative workarounds to their restrictive UI. Before long it was clear that, however far Threadbare had come, the project could never reach its potential while hosted on Discord. On top of that, the name "Threadbare" was already hotly contested among fashion businesses, so a branding headache was brewing too.

Those problems led us to become Cut Thwice in 2024, adopting the name of a resale store I'd cooked up in in PHP back in 2014 (the earliest Internet Archive scrape, unfortunately, only goes back to 2018) in the University of Auckland's Math Department computer labs. It was entirely janky, but for whatever reason I kept paying the server costs for the next ten years, and I still have the domain today. CutThwice felt like a natural transition, and a part of me was excited to return to a cherished project, but with a more ambitious aim. There were some minor tweaks to the Discord technology, but the biggest changes here were building a standalone website system and a greater commitment to making the most comprehensive secondhand designer search platform around. The branding overall left a lot to be desired, but the real culprit was a particular shade of purple and a logo that looked like a paint splat with "Cut Thwice" etched into it; a warped page torn from the Nickelodeon brand book. Nevertheless, we pressed on, and this became the project's turning point, the shift from a humble experimental side project to a deliberate attempt at building something that could meaningfully shift consumer behavior in the secondhand designer market. This showed up immediately as far more labor and skill poured into building a fully-fledged standalone search platform, alongside a sharp increase in vendor and brand coverage. That surge in development came in no small part from the relief of no longer being chained to Discord, which let me build a technology tailor-made for the secondhand designer community to my heart's content. We quickly built a fully searchable database with much more comprehensive features that let users efficiently scour the entire market. On top of that, we added more depth to the experience by scaffolding different search practices; compartmentalizing the market and layering in filtering to navigate it. While the project was still focused on refining the technical elements, the cultural and creative grooves in our heads began to deepen as we started envisioning, more ambitiously, what the project could become, reigniting my passion for secondhand designer culture through the act of building an innovative platform for consumers like myself. Little did I know this would just be the kindling of what KURB is today.

After developing Cut Thwice for almost a year, it was high time for our most important evolution: in early 2025, we rebranded the whole project as KURB. As the project gained momentum in the community, it only made sense to focus on building a brand we could happily live in for the foreseeable future. The branding took months and plenty of heated internal debate, aligning the technology with SEO opportunities, communication shifts, and countless other moving parts. But it was the perfect time to make the change; the opportunities for the project were sprawling in every direction, and having the right vehicle to explore them mattered more than ever. KURB signaled a whole new dynamic for how we'd exist within the secondhand designer ecosystem, far more than a new coat of paint. We started trialing more ambitious creative content, collaborating with prominent figures in the ecosystem, publishing blog posts that dug into the nuances of the secondhand designer market, running tighter iterative rounds of feedback, building analytical content from our database, and finding new ways to engage with the secondhand designer landscape at large.

That didn't mean we slacked on the technology front, if anything, we quadrupled down. We trained our own AI classification system to improve listing quality and search notifications, added reverse-image search against our database, built out more advanced filter interactions (saved filters, keyword search, filter exclusion, resumed searches), shipped price-drop notifications, overhauled the UI and UX, integrated proxy information, added on-the-fly translations, deal evaluations, and favorite collections, and as I write this, I'm still tuning our color-prediction and price-intelligence models. All the while, we kept refining our core technology: several revisions of the KURB search stack, a huge improvement to sold-item checking, and a massive expansion in vendor and brand coverage (we're now at almost 900 stores and over 500 brands). Don't just take my word for it, our weekly tech stories and monthly newsletters are a running record of how constantly KURB keeps improving. Despite all this, one thing becomes clearer every day: we're just scratching the surface of what KURB is going to become.

KURB has gone through many twists, and I can guarantee there will be more to come. What's remained true is our steadfast desire, against all odds, to make the most comprehensive secondhand designer search platform, solely focused on getting you everything you need to acquire the clothing of your dreams (something my 16-year-old self wished for between 3 a.m. Yahoo Auction bid wars and WC3-DotA games). What KURB becomes next is still unknown though I'm confident you'll love it as much as we do.